This project is intended to test both subjective (self-report) and objective (chemical testing) methods of assessing the prevalence of workforce drug abuse. The research will contrast several subjective reports of drug use with objective assessments based on urinalysis, employing the cooperation of a large company (Westinghouse with its employees and applicants as subjects) to develop valid, effective methods and guidelines for estimating drug and alcohol abuse prevalence in industry. This research meets a critical NIDA requirement (re: Announcement No. DA-87-26) "...to develop and utilize valid and reliable methods of assessing incidence and prevalence of drug use at the workplace ... where possible, drug usage should be documented by appropriate measures of levels in body fluids." The specific objectives of the research are: to test subjective (self- report) methods of assessing substance abuse among job applicants and in-service employees (methods to be tested include self-administered questionnaires (SAQ) and personal interviews, collected both on-site and off-site, and telephone interviews); to validate subjective assessment techniques against the objective criterion of chemical testing; to determine the relationship between the dichotomous measure of drug use produced by chemical testing and the more detailed descriptions of drug and alcohol use patterns and circumstances; to produce accurate estimates of the prevalence of drug abuse among employees at a major worksite; to provide a data base on employee substance abuse that can be used for future research studies; and to develop valid, effective methods and guidelines for use in estimating drug and alcohol abuse prevalence in industry. The overall research design combines a classic criterion validity design (with urinalysis results as the criterion) with a rigorous experimental contrast the subjective assessment methods. Data is to be collected on a total of 1800 workers (900 job applicants and 900 in-service employees). The 900 job applicants will be sampled exhaustively; the in-service employees will be randomly sampled from the workforce at three sites; a 25% radom sample will provide the sample sizes required at each site. The self-report data and the urinalysis data will be subjected to several types of analysis.